Patients and methods. The inclusion criteria in this retrospective study were: patients aged over 65 at the time of diagnosis, presenting with extensive eczematiform eruption (> 20 P. loo of body surface) and lasting for more than one month. Eczema on stasis dermatitis of the lower limbs and generalized contact eczema were excluded. Eighty-three patients followed between January 1990 and January 1999 were included. The clinical, biological, histological and evolving characteristics were analyzed.

Results. Mean age of patients was 77 +/- 8 years and the male female sex ratio was 2.4. Patients received a mean of 4.0 +/- 2.6 drugs/patient, consisting essentially of cardiovascular and psychotropic agents. The cutaneous eruption had evolved a mean if 12.5 months (1 to 48 months) before diagnosis. Eczema was pruriginous in 92 P. 100 Of cases. Frequent relapses were observed in 68 P. 100 of cases. Precise etiologic diagnosis was retained in 48 patients (58 P. loo). This was disseminated contact eczema (n = 19), lymphoma cutis (n = 10), atopic eczema (n = 7), scabies acariasis (n = 6) and pemphigoid (n = 6). No etiologic diagnosis was retained in the remaining 35 patients (42 P. 100). Comparison of the characteristics in the 2 groups showed excessive consummation of medicinal products (P = 0.024), predominant eruption of sun-exposed areas (P = 0.004) and a greater frequency of histological images of keratinocyte necrosis (P = 0.0072) in patients presenting eczematiform eruptions of unknown etiology.

Discussion. These observations suggest the eventual responsibility of medicinal products in the occurrence of certain extensive and chronic eczematiform eruptions in the elderly. However, the delays of imputability of various causal drugs were often longer than those currently admitted for toxidermia, and the withdrawal of potentially imputable agents rarely led to spectacular improvement in the lesions. A case test report is in progress to specify this hypothesis.